Research On The Modeling And Analysis Of Intensity-modulated Optical Fiber Sensor And Its Implementation

Compared with the traditional sensor,fiber sensor have advantages of compact structure, flexible design, immune to electromagnetic interference, resistance to harsh environments, etc. and it can achieve high accuracy, stability and sensitivity measurements. Intensity-modulated optical fiber sensor is based on the fluctuations of the light intensity to achieve the measurements.The model of the intensity-modulated optical fiber sensor is based on the near-infrared absorption theory, and the depth research have been done about light intensity compensation, reflective probe, weak signal processing in order to complete development of the sensor and experimental research.Firstly, by studying of the spectrum absorption and measurement methods, classifying different forms of objects (gas, liquid, solid) measurement principles, the two models based on transmission and diffuse reflection are established. Transmission model is the harmonic detection differential mathematical model, based on the Lambert-Beer law. The model shows the relationship between gas concentration and the first harmonic, the second harmonic are the linear relationship. Diffuse reflection detection is aim at the measurement of solid and opacity liquid, and which is build on the K-M law,from the mathematical model, know that the concentration of the object have no descriptive liner relationship with the output, so introduce a aritificial neural network and build the measurement model of the neural network in the background of diffuse principle.According to the theoretical analysis and modeling, intensity-modulated optical fiber sensor system is designed and developed. In light path design of the sensor system, in order to improve the precision and stability of the sensor, the light intensity compensation based on the neural networks and a combined optical fiber bundle probe probe are proposed after analyzing the influence of various factors and the performances of the sensor. Their optimizing design methods based on theoretical models are presented according to the theoretical modeling and simulation analysis.The photoelectric signal detecting system is an integral part of the sensor. The pre-amplifier circuit and weak signal processing circuit are designed and implemented according to the principle of photoelectric signal detection.Based on the theories of the correlation detection and the characteristics of the optical fiber sensor, an analog signal processing system and a digital signal processing system used for weak signal detection are designed. Through the analysis and the test, it turns out that the digital signal processing system is more reliability, stability and flexibility than the analog signal processing system.Finally, Intensity-modulated optical fiber sensor based on transmission detection and diffuse reflectance detection are developed. The experiment, through which the carbon monoxide concentrations are detected to verify the mathematical model of the differential harmonic detection, has a result that shows a good linear relationship between the carbon monoxide concentration and the first harmonic amplitude. Although the experiment that detects the water in the paper based on the diffuse reflection principles has a result shows the nonlinear relationship between concentration and output, a good detection effect is got by using the nonlinear correction of the neural network.